Nvidia advances its memory-chip alliance strategy alongside the rollout of the Vera Rubin architecture to bolster next-generation accelerator supply chains.

Nvidia’s shares rebounded 8.6 percent following its quarterly earnings report, closing at €195.60 by Thursday and ending its longest losing streak since 2022. The initial after-hours reaction had been muted, reflecting stratospheric market expectations, before the stock snapped back with force. The recovery is driven not only by the earnings print but by strategic developments across Nvidia’s supply chain and product roadmap, signaling an acceleration in the AI infrastructure buildout.

For fiscal second quarter, Nvidia reported revenue of $96.2 billion, a 106 percent increase from $46.7 billion a year earlier. Earnings per share nearly doubled to $2.22, while gross margin expanded to 75.0 percent. The data center segment, which now constitutes 92 percent of total revenue, generated approximately $89 billion. This figure comfortably surpassed analyst consensus of $86.33 billion and represents a 117 percent year-over-year jump. Looking ahead, CEO Jensen Huang guided to $108 billion in revenue for the current quarter, exceeding consensus estimates and notably excluding any China data center revenue. On the earnings call, Huang stated directly: "AI has reached its tipping point. It does useful work. Its tokens are productive and profitable. Now compute equals revenue."

Central to Nvidia’s forward momentum is a multi-year partnership with SK Hynix, confirmed this week, to secure AI memory chips for next-generation hardware. The agreement covers the upcoming Vera Rubin systems, the Vera CPU, RTX Spark, and Jetson Thor. Nvidia has previously warned of "extreme pricing conditions" in the memory market, and this tie-up functions as a strategic hedge against supply chain bottlenecks that could otherwise throttle growth. To address rising memory costs, the company is also deepening its ecosystem partnerships with Marvell, MediaTek, and Qualcomm, marking a deliberate shift toward embedding itself across the semiconductor value chain rather than operating in isolation.

Alongside the memory alliance, Nvidia unveiled two technologies designed to enhance future accelerator performance: NVLink Fusion and NVHBM architectures. These innovations promise 30 percent higher memory bandwidth and 15 percent lower power consumption compared to HBM4e, while freeing up 25 percent more die area for compute logic. On the networking front, the company introduced the Mellanox Switch 920-9B210-00FN-0D0, engineered for large-scale AI clusters. Featuring 64 ports operating at 400 gigabits per second each under the NDR standard, the switch delivers a total switching capacity of 51.2 terabits per second. Its integrated SHARPv3 technology reduces latency in collective communication by 40 percent—a critical advantage for data center operators relying on near-instantaneous GPU data exchange.

The Vera Rubin platform, Nvidia’s next major growth driver, has entered full production and is projected to account for roughly one-fifth of data center revenue in the current quarter. Industry reports indicate that the supply chain for this generation is twice the size of that built for Grace Blackwell, underscoring robust demand expectations. This optimism is reinforced by broader market commitments: the five largest hyperscalers are expected to increase capital expenditures from $800 billion this year to $1.3 trillion next year, and Amazon Web Services has committed to purchasing two million Nvidia GPUs alongside the new Vera CPU platform.

From a technical perspective, Nvidia’s current price of €195.60 sits 3.4 percent below its 52-week high of €202.50. More significantly, it trades 16 percent above its 200-day moving average of €168.64, highlighting how rapidly the uptrend has extended. The pivotal question for investors is whether the company’s expanded supply chain capacity will successfully convert into sustained revenue over the coming quarters. While Nvidia has flagged rising memory costs as a potential headwind, the SK Hynix partnership and broader ecosystem integrations are direct mitigants.

Ultimately, Nvidia is transitioning beyond a pure GPU vendor. Through the Vera Rubin rollout, advanced networking infrastructure, and fortified memory partnerships, the company is positioning itself to control the entire AI infrastructure stack—from storage and switching to core compute engines. Whether this vertical integration will deliver returns at the pace the market anticipates remains to be seen, but the foundational components are firmly in place.